Description
Pricing Details
The pricing model is based on a pay-as-you-go system, along with the option to reserve resources or use on-demand services; there is no free plan explicitly mentioned in the available content. For cloud GPUs, prices vary depending on the type of processor used, such as the RTX 4090 at approximately $0.59 per hour, the RTX 6000 Ada at approximately $0.77 per hour, and the L40S at approximately $0.86 per hour, while costs rise for more powerful processors like the A100, H100, and H200, ranging from roughly $1.39 to $5.49 per hour or more; some models, such as the B200, may reach the same level or higher. The Serverless system, on the other hand, is based on per-second billing, where the cost of Flex Workers starts at around $0.00240 per second, with Active Workers available at a lower cost that can offer a discount of approximately 30%, and prices vary depending on the type of GPU used. With the Instant Clusters service, you can run multiple GPU clusters of up to 64 units on a pay-as-you-go basis. Prices start at around $1.79 per hour for the A100 SXM, while more powerful models like the H100 SXM and B200 require contacting the sales team to determine pricing. The platform also includes a storage system ranging from $0.05 to $0.14 per gigabyte per month, with options such as Container Disk, Volume Disk, and Network Storage. Additionally, there are paid APIs (Public Endpoints) that include voice models starting at $0.05 per 1,000 characters, image models ranging from $0.005 to $0.14 per request, text models starting at $0.00001 per million tokens, and video models at approximately $0.12 per 5 seconds or depending on the model used.